Following is the section of a White House budget book addressing the issue of differing results from comparisons of federal and private sector pay.

Federal worker compensation receives a great deal of attention, particularly in comparison to that of private sector workers. Comparisons of the pay and benefits of Federal employees and private sector employees must account for factors affecting pay, such as differences in skill levels, complexity of work, scope of responsibility, size of the organization, location, experience level, and exposure to personal danger, and should account for all types of compensation, including pay and bonuses, health benefits, retirement benefits, flexibility of work schedules, job security, training opportunities, and profit sharing.
A series of reports released in January 2012 by the Congressional Budget Office (CBO) accounted for some, but not all, of the factors described above. CBO found that prior to the three-year Federal pay freeze, Federal pay, on average, was slightly higher (2.0 percent) than comparable ble private sector pay. CBO reported that overall Federal sector compensation (including benefits) was on average substantially higher, but noted that its findings about comparative benefits relied on far more assumptions and were less definitive than its pay findings. The CBO study also excluded forms of compensation, such as job security, that favor the Federal sector, and factors such as training opportunities and profit sharing that favor the private sector.
Perhaps more importantly, CBO emphasized that focusing on averages is misleading, because the Federal/ private sector differentials vary dramatically by education and complexity of job. Compensation for highly educated Federal workers (or those in more complex jobs) is lower than for comparable workers in the private sector, whereas CBO found the opposite for less educated workers. These findings suggest that across-the-board compensation increases or cuts may not be the most efficient use of Federal resources.
The CBO reports focus on workers and ask what employees with the educational backgrounds and other characteristics of Federal workers earn in the private sector. An alternative approach, used by the Federal Salary Council, focuses on jobs and asks what the private sector would pay people with the same roles and responsibilities as Federal workers. Unlike CBO, which found that Federal pay is (on average) roughly in line with private sector pay, the Federal Salary Council found that in 2014 Federal jobs paid 35 percent less than comparable non- Federal jobs.
There are possible explanations for the discrepancy in the CBO versus the Federal Salary Council findings. First, methodological issues around the classification of Federal and private sector jobs introduce considerable uncertainty into the Federal Salary Council approach. It is significantly easier to compare college graduates in Federal versus private sector jobs than it is to determine what private sector job is most comparable to a given Federal job. Second, the studies ask fundamentally different questions, so their different answers are not necessarily in conflict. It could be the case that Federal and private sector workers with similar characteristics are paid about the same, but that jobs in the Federal sector are underpaid relative to their private sector counterparts. That would imply that, at least in some jobs, the Federal government could have difficulty hiring and retaining workers with the same skills or managerial experience as their counterparts in equivalent private sector jobs This could be a reason for concern, given the decline in the size of the Federal workforce relative to the population and the increasingly supervisory role it plays (e.g., supervising contractors and State and local governments).